A process for upgrading a mirrored shared-nothing database system comprises a sequence of short well-defined idempotent steps, and at least one non-idempotent step involving transforming a master catalog. The upgrade process is managed and controlled by a state machine that has a persistent memory running on the master node. In the event of a failure or crash during an idempotent step, the process stops the database in the current state and repeats the step. If a failure or crash occurs during a non-idempotent step, the upgrade process is rolled back to the beginning and repeated.
Legal claims defining the scope of protection, as filed with the USPTO.
1. A method for upgrading a mirrored database, comprising: running an upgrade process comprising a sequence of well-defined idempotent steps and a non-idempotent step on the database, each step having as possible outcomes a success, a failure, or a crash; monitoring the status of the upgrade process and the database state using a state machine having a persistent memory; upon an error occurring during one of said idempotent steps of the upgrade process, stopping the upgrade process in a current state of the database and repeating said one idempotent step; and upon an error occurring during said non-idempotent step, rolling back said upgrade process to a beginning and repeating the upgrade process.
2. The method of claim 1 , wherein said mirrored database comprises a master node and a plurality of segment nodes, said segment nodes having one or more pairs of primary and mirror databases, and wherein said state machine is run on said master node.
3. The method of claim 2 , wherein said non-idempotent step comprises transforming a catalog on said master node.
4. The method of claim 3 further comprising aborting said upgrade process upon the occurrence of a failure.
5. The method of claim 3 , wherein said upgrade process comprises propagating said transformed catalog to said primary and mirror databases.
6. The method of claim 2 , wherein said upgrade process comprises turning off and preserving a mirror database as a backup for failure of a primary database.
7. The method of claim 5 , wherein said upgrade process further comprises recovering the primary database from said mirror database in the event of a failure.
8. Computer readable non-transitory storage medium for storing instructions for controlling the operation of a computer for upgrade of a mirrored database system, comprising instructions for: running an upgrade process comprising a sequence of well-defined idempotent steps and a non-idempotent step on the database, each step having as possible outcomes a success, a failure, or a crash; monitoring the status of the upgrade process and the database state using a state machine having a persistent memory; upon an error occurring during one of said idempotent steps of the upgrade process, stopping the upgrade process in a current state of the database and repeating said one idempotent step; and upon an error occurring during said non-idempotent step, rolling back said upgrade process to a beginning and repeating the upgrade process.
9. The computer readable medium of claim 8 , wherein said mirrored database comprises a master node and a plurality of segment nodes, said segment nodes having one or more pairs of primary and mirror databases, and wherein said state machine is run on said master node.
10. The computer readable medium of claim 9 , wherein said non-idempotent step comprises instructions for transforming a catalog on said master node.
11. The computer readable medium of claim 10 further comprising instructions for aborting said upgrade process upon the occurrence of a failure.
12. The computer readable medium of claim 10 , wherein said instructions comprise instructions for propagating said transformed catalog to said primary and mirror databases.
13. The computer readable medium of claim 9 , wherein said instructions comprise instructions for turning off and preserving a mirror database as a backup for failure of a primary database.
14. The computer readable medium of claim 13 , wherein said instructions further comprise instructions for recovering the primary database from said mirror database in the event of a failure.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
February 10, 2012
June 3, 2014
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.